Efavirenz (EFV) is a non-nucleoside reverse transcriptase inhibitor (NNRTI) with a long half-life, allowing for once-daily dosing. Although it is generally well-tolerated and widely used, a major disadvantage of EFV is its relatively low barrier to resistance. The single K103N mutation confers high-level resistance to EFV, as well as to all other NNRTI!?s. Acquisition of HIV-1 resistance is likely to result from repeated exposure of the virus to subtherapeutic drug concentrations. When terminating EFV therapy it is recommended that EFV be stopped 1-2 weeks earlier than concomitant antiretrovirals (ARVs) to account for its long washout period. Alternatively, substituting EFV with a protease inhibitor for 1 to 4 weeks has also been recommended. Both strategies attempt to avoid prolonged EFV exposure following discontinuation of concomitant ARVs, thereby lowering the risk of acquiring new resistance mutations. Though widely utilized, neither method has been validated by pharmacokinetic or clinical studies. Genetic differences in CYP2B6 have been shown to contribute greatly to observed variability in EFV clearance; hence, inter-patient genetic differences may significantly influence the manner in which efavirenz should be discontinued in order to reduce the risk of acquiring HIV-1 resistance mutations. Multiple polymorphisms in CYP2B6 have been identified. Substitutions at positions 516, 785, and 1459 are the most common variants, and appear to be associated with alterations in CYP2B6 activity. We propose a 2-part study. Part I is to be conducted in HIV-negative volunteers, and consists of: (1) determination of EFV pharmacokinetic (PK) values (following 13-15 days of EFV 600 mg/day) for comparison between CYP2B6 genotypes (including wild-type and allelic variants at positions 516, 785 and 1459), (2) measurement of single EFV plasma concentrations 7, 14, 21, and 28 days after the drug is stopped to assess for differences in EFV exposure between CYP2B6 genotypes, and (3) comparison of bupropion PK values following administration of a single bupropion dose (a) prior to EFV administration and (b) after 13-15 days of EFV, to evaluate whether EFV modulates CYP2B6 metabolism using bupropion as a probe for CYP2B6 metabolic activity. Part II is to be conduced in HIV-positive patients who are planning to discontinue (d/c) their EFV therapy, and includes: (1) determination of EFV PK values prior to d/c of therapy for comparison between different CYP2B6 genotypes, (2) measurement of single EFV plasma concentrations 7, 14, 21 and 28 days post-EFV to assess for differences in exposure across CYP2B6 genotypes, and (3) genotypic evaluation of HIV-1 drug resistance performed in any patient with an HIV RNA !Y 1,000 copies/mL, prior to and after EFV discontinuation (up to 12 months after stopping EFV), as a preliminary study to evaluate the incidence of EFV resistance mutations based on CYP2B6 genotype. Data from this study will help to describe the relationship between CYP2B6 genetic polymorphisms and EFV disposition, specifically addressing the effect of CYP2B6 genotype on prolonged EFV exposure following discontinuation of therapy. In addition, this study will determine whether EFV modulates CYP2B6 activity as characterized by bupropion phenotyping. Lastly, the investigation will attempt to provide preliminary information about whether genetic variability in CYP2B6 is related to the incidence of NNRTI-resistance after stopping or changing EFV therapy.